The demonstration of lysosomes by the controlled temperature freezing-sectioning method

The distribution of acid phosphatase in rat liver homogenates is almost exclusively cytoplasmic. Palade (1951) and Berthet and de Duve (1951) found that 95% of the total activity was in the cytoplasmic fraction, the nuclear activity being regarded as due to contamination. However, the histochemical localization of acid phosphatase activity by the lead phosphate method (Gomori, 1941) conflicted with these results and showed, in the liver cells, a diffuse distribution with activity either in the cytoplasm or in the nucleus or in both.

De Duve (e.g. 1959; see also Gianetto and de Duve, 1955) described a distinct group of cytoplasmic particles called ‘lysosomes’ which contained a number of acid lytic enzymes including acid phosphatase. These enzymes are contained within the lysosomes by a lipoprotein membrane and are released when this membrane is damaged.

Holt (1959) examined the lead phosphate procedure critically, and by modifying both the method by which the tissue was prepared and also the constitution of the incubation medium, was able to demonstrate structures in sections of rat liver and kidney which resembled the lysosomes of tissue fractions. Although his techniques did not allow him to show that these were inert particles which became activated only when the membrane was modified, yet he was able to enhance their activity and even to release the acid phosphatase completely by procedures similar to those used by de Duve.

Thus the biochemical data show the lysosome as an inert particle which cannot be activated until some damage is inflicted on the bounding membrane. Only after this membrane has become altered can enzymic activity be demon-strated inside the particles themselves (Holt, 1959). The reports of lysosomal enzymes, such as acid phosphatase, occurring diffusely throughout the cell, or in nuclei, would seem to be due either to further disruption of the lysosomes, with subsequent diffusion of the enzyme itself, or to diffusion artifacts of the histochemical method.

Thus the complete identification of particles in sections with lysosomes is made unreliable by various types of diffusion artifact. For example, the presence of activity on small cytoplasmic grains might represent only sites of adsorption of either the enzymes or the final dye. To prove conclusively that particular types of granules correspond to the biochemists’ lysosomes requires, ideally, that they should be shown to be (1) inert when intact; (2) activated by treatments that will damage membranes. It might be regarded as more satisfying if in addition it could be shown that the acid phosphatase of these granules can be made to diffuse throughout the cell, and so give rise to the earlier histochemical localization of this enzyme. This might be advantageous, since it could otherwise be argued that the acid phosphatase demonstrated on the granules represented only a very small, and perhaps even an unimportant, fraction of the total cellular content of this enzyme.

Of these three criteria, only the last two have up till now been demonstrated, owing to the fact that all the treatments that were required for the production of suitable sections have been damaging to membranes (Holt, 1959). With the advent of the controlled temperature freezing-sectioning method, which appears to preserve lipid-protein complexes (Chayen and others, 1961), it seemed advisable to test whether a complete cytochemical identification of the particles with the biochemists’ lysosomes could be achieved.

Albino Wistar male rats, fed on a complete diet (MRC rat diet B41), were killed by placing them under a funnel through which nitrogen was passed from a cylinder at a rate of over 1 1 per min.

A sample cube of side approximately 5 mm was taken from the liver and frozen immediately in a glass tube which had been previously cooled in solid carbon dioxide ice. Sections were cut at 8 μ on a freezing cryostat microtome at about —25° C, with the knife cooled to about — 70° C.

Sections were treated by the Gomori acid phosphatase procedure as modified by Holt (1959); the incubation time in the substrate varied from 5 to 60 min. Serial sections were treated for 5 min with either a 10% solution of neutralized formalin (40% formaldehyde) containing 0·9% sodium chloride, or an 0·25% solution of ‘triton X–100’, or an 0·25 M solution of sucrose. The sections were washed in running water and then incubated as described above. The staining procedure was controlled by the addition of 0·01 M sodium fluoride to the incubation medium to inhibit the enzyme.

The results are summarized in table 1 (see Appendix, p. 209).

Appendix

No stain was discernible in the sections incubated for 5 or 10 min (fig. 1, a), whereas when the incubation time was extended to 20 min, small black granular structures could be seen in some of the liver cells (fig. 1, b). Larger stained structures distributed at the cellular periphery appeared in most cells after incubation for 40 min, while prolonged incubation (60 min) resulted in very diffuse cytoplasmic and nuclear staining (fig-I,c).

No stain was observed in the sections incubated for 5 min, but incubation for 10 min revealed small black granules in some of the liver cells (fig. 1, D), similar to those seen in the untreated sections which had been incubated for 20 min. Most of the liver cells contained blackened cytoplasmic granules and there was some nuclear reaction in the sections treated for 20 min (fig. 1, E). More prolonged incubation resulted in diffuse cytoplasmic and intense nuclear staining (fig. 1, f).

Black granules appeared in some of the liver cells after incubation for only 5 min. The sections treated for 10 min showed stained granules in most cells and there was some diffuse cytoplasmic staining. Incubation for longer periods resulted in diffuse cytoplasmic and intense nuclear staining.

No deposition of lead sulphide was detected in these sections even after prolonged incubation in the substrate.

It has been shown by de Duve and his associates (e.g. 1959) that glycerophosphate does not penetrate the membrane of intact lysosomes which have been isolated by homogenization and differential centrifugation. Hence the activity of such particles is demonstrated, biochemically, after the enzymes have been rendered soluble; to achieve this it is necessary to disrupt the bounding membrane. Holt (1959) postulated that formalin fixation may partially denature or modify the lysosomal membrane and allow easier access of the substrate to the enzyme, without permitting the latter to diffuse out of the particle. This could not be proved by the use of tissue fixed in formalin, but only by the application of a technique which did not involve chemical fixation of the tissue. In this study, sections prepared by the controlled temperature freezing-sectioning method and treated with formalin showed stained granules after short incubation, whilst untreated sections remained unstained. Thus the lysosomes had been activated by formalin treatment, but remained inert in the untreated section.

Activated lysosomes were observed in the untreated sections, but only after more prolonged incubation. This was probably due to the effect of heat on the lysosomal membrane and is in agreement with the finding that intact, inactive lysosomes in tissue homogenates are activated by incubation at 37° C at pH 5 (de Duve, 1959). Short periods of incubation allow access of the substrate to the enzyme, while longer periods result in diffusion of the enzyme with diffuse cytoplasmic and nuclear staining.

‘Triton X-100’ releases acid phosphatase from liver lysosomes (Gianetto and de Duve, 1955). Similar results were obtained with sections prepared by the controlled temperature freezing-sectioning method. No staining reaction occurred even after prolonged incubation in the substrate. It would seem that the enzyme was released from the lysosomes and had diffused out of the section.

Lysosomes in tissue homogenates are activated by repeated freezing and thawing (de Duve, 1959). Holt (1959) found that particles stained very rapidly in frozen-dried sections and he suggested that some modification of the lysosomal membrane had occurred. The controlled temperature freezing-sectioning method applied in this study does not appear to have had the same effect as freezing-drying on the lysosomal membrane, since the integrity of the membrane seems to be preserved. This is shown by the fact that the lysosomes were inactive until subjected to damaging agents like heat, formalin, and ‘triton X–100’. With increasing severity of the treatment, activity could be demonstrated either in the lysosomes alone, or in the lysosomes and cytoplasm, or in the cytoplasm and nucleus (see table 1). The most severe treatment (triton X–100) released the enzyme and no activity could be demonstrated.

Thus the three criteria that are required to prove that granules correspond to the biochemists’ lysosomes have been fulfilled in this study. The granules have been shown to be inert when intact; they have been activated by treatments that will damage membranes; and the acid phosphatase of the granules has been made to diffuse throughout the cell and give rise to intense but diffuse cytoplasmic and nuclear activity.

I wish to acknowledge my gratitude to Professor G. J. Cunningham for his encouragement and to Dr. S. J. Holt for his guidance in this work. I am also much indebted to Dr. J. Chayen for most stimulating and helpful discussion. I should like to thank Mr. A. A. Silcox for his skilful assistance and Mr. A. L. E. Barron for the photomicrography.

I am indebted to the British Empire Cancer Campaign for financial assistance and I wish to thank the Trustees of the Prophit Fund for a research studentship.